Processing a point cloud
At least one embodiment relates to a method for smoothing (filtering) the geometry of a point cloud frame by performing an analysis and filtering of said geometry point cloud in a 2D space, without reconstruction of 3D samples in a 3D space, and by using a flexible filtering operator that, in addition to moving existing points, can also remove points or add new ones.
1. A method comprising:
obtaining a patch representative of a geometry of 3D samples of a point cloud frame projected on a first projection plane,
obtaining at least one neighboring patch of the patch from at least one candidate patch, wherein a candidate patch is a neighboring patch when a 3D bounding box of the candidate patch intersects a 3D bounding box of the patch,
obtaining a reference contour by re-projecting at least one 2D sample of the at least one neighboring patch of the patch onto the first projection plane, and
modifying spatially at least one 2D sample of the patch to fit the reference contour.
2. The method of claim 1 , wherein the 2D samples of the patch forms a contour of said patch.
3. The method of claim 2 , wherein the contour of the patch is modified by adding and/or removing at least one 2D sample of the patch.
4. The method of claim 3 , wherein a 2D sample of the contour of the patch is modified if it is either inside or outside the reference contour.
5. The method of claim 2 , wherein the contour of the patch is modified by relocating at least one 2D sample of the patch along a projection axis of the patch.
6. The method of claim 2 , wherein the contour of the patch is modified by adding at least one additional 3D sample between at least one 3D sample corresponding to a 2D sample of the patch and at least one 3D sample corresponding to a 2D sample of the reference contour.
7. The method of claim 1 further comprising filtering attributes of said 2D samples of the patch.
8. The method of claim 7 , wherein attributes of said 2D samples of the patch comprises color information or depth information.
9. The method of claim 1 , wherein re-projecting at least one 2D sample of the at least one neighboring patch of the patch onto the first projection plane comprises:
obtaining 3D samples by inverse-projecting 2D samples of the at least one neighboring patch according to a projection axis associated with the neighboring patch,
projecting the obtained 3D samples along a projection axis associated with the patch onto the first projection plane.
10. The method of claim 1 , wherein the at least one 2D sample of the at least one neighboring patch is a 2D sample located at a boundary of the at least one neighboring patch.
11. The method of claim 1 , wherein only one 2D sample re-projected onto the first projection plane is kept per location a canvas of the patch.
12. The method of claim 11 , wherein the only one re-projected 2D sample that is kept for a given location is a re-projected 2D sample having a depth that is closest to a depth of a 2D sample of the patch located at the same location.
13. An apparatus comprising one or more processors configured to:
obtain a patch representative of the geometry of 3D samples of a point cloud frame projected on a first projection plane;
obtain at least one neighboring patch of the patch from at least one candidate patch, wherein a candidate patch is a neighboring patch when a 3D bounding box of the candidate patch intersects a 3D bounding box of the patch;
obtain a reference contour by re-projecting at least one 2D sample of the at least one neighboring patch of the patch onto the first projection plane; and
modify spatially at least one 2D sample of the patch to fit the reference contour.
14. The apparatus of claim 13 , wherein the at least one 2D sample of the at least one neighboring patch is a 2D sample located at a boundary of the at least one neighboring patch.
15. The apparatus of claim 13 , wherein the 2D samples of the patch forms a contour of said patch.
16. The apparatus of claim 15 , wherein the contour of the patch is modified by adding and/or removing at least one 2D sample of the patch.
17. The apparatus of claim 13 , wherein attributes of said 2D samples of the patch comprises color information or depth information.
18. A non-transitory computer-readable medium including instructions for causing one or more processors to perform a method according to claim 1 .
19. The apparatus of claim 13 , wherein re-projecting at least one 2D sample of the at least one neighboring patch of the patch onto the first projection plane comprises:
obtaining 3D samples by inverse-projecting 2D samples of the at least one neighboring patch according to a projection axis associated with the neighboring patch,
projecting the obtained 3D samples along a projection axis associated with the patch onto the first projection plane.